Xponent Dataset Projection Methodology

Glad to see the response for my last post based on Xponent Dataset.😃 Here further elaborating the Xponent Dataset Projection Methodology in detail -

Xponent data is more of projection prescription data i.e. projecting prescription volume data for non-panel stores using actual data of panel stores. It uses Geospatial statistics technique that implies closer stores tend to have same prescribers with similar prescription volume and product mix.

To determine the prescription volumes for non-panel stores, we consider 3 factors -

1. Number of panel stores near non-panel stores (min-2 and max-10)

2. Distance between panel and non-panel stores. Distance should be less for better estimation

3. Store size. Applying weightage methodology to normalize store size between panel and non-panel stores

Lets take an example - Suppose actual prescriptions (Rx) for Dr.M be 100 Rx and we need to project Rx for Dr.M. Below are the steps for projecting data -

Assumptions -

1. Suppose there are 2 non-panel stores near Dr.M

2. There are 3 panel stores closer to non-panel stores

Steps to project Rx -

1. Capture the actual known data for all the panel stores near Dr.M. Captured Rx for panel stores = 20 + 30 + 20 = 70 Rx

2. Using geospatial projection methodology, try to create factors for both non-panel stores based on the distance and size of panel stores

3. Suppose for non-panel store 1, factors from each panel store are 0.2, 0.4 and 0.24 respectively

4. Estimated Rx for non-panel store 1 = (20*0.2) + (30*0.4) + (20*0.24) = 20.8 Rx 

5. Similarly, for non-panel store 2, factors are as follow - 0.1, 0.2 and 0.3 respectively

6. Estimated Rx for non-panel store 2 = (20*0.1) + (30*0.2) +(20*0.3) = 14 Rx

7. Total Rx for Dr.M = Actual captured Rx from 3 panel stores + Estimated Rx for non-panel store 1 + Estimated Rx for non-panel store 2

8. Total Rx = 20 + 30 +20 +20.8 +14 = 104.8 Rx

Hope it is helpful! Drop any questions or insights in comment. 😊 

hashtag#pharma hashtag#pharmaindustry hashtag#healthcare hashtag#pharmaconsulting hashtag#dataset hashtag#iqvia hashtag#healthcareconsulting hashtag#xponentdataset hashtag#prescriptions hashtag#xponent